Genetics of Celiac Disease

The Genetics of Celiac Disease (1)Edited: This was originally posted as two entries, now combined to have all the information in one spot.  

One of my main reasons for getting 23andMe testing done was to figure out if I could possibly have celiac disease.  I went gluten-free about five years ago and have felt so much better that I didn’t want to go back to eating gluten in order to get the testing done for celiac. You see, the ‘Catch-22’ of getting diagnosed for celiac disease is that you have to be eating gluten for a while before getting tested.   So I was hoping that the genetic testing would rule out the possibility of celiac for me (and my family).

First a little background information...
Celiac disease (spelled coeliac in Britain) is an autoimmune disorder in which gluten causes damage to the villi in the small intestines.  Gluten is a protein found in wheat, barley, rye, and spelt. There are many resources on the internet with great information about celiac disease and the symptoms involved. is a good starting point for anyone wanting to learn more.

Genetic component to Celiac…
There isn’t one gene polymorphism that causes everyone who has it to get celiac disease, but there are polymorphisms that must be present in order for a person to be susceptible to celiac disease.  So if the polymorphism isn’t present, you can nearly always rule out the disease.

Virtually everyone with celiac disease has either the HLA-DQ2 or HLA-DQ8 alleles. Specifically,  90-95% of people with celiac disease have HLA-DQ2.5, and 5 – 10% of people have HLA DQ8. (Some studies also list HLA 2.2 as a possibility for celiac as well.) Almost 25% of the population has HLA-DQ2.5, and that percentage grows to about 30% when adding in HLA DQ8. [ref]  There are a couple of single nucleotide polymorphisms (SNPs) that you can look at (listed below) to determine if you have those HLA types.   With only about 1% of the population having celiac disease, you can see that just having the HLA type doesn’t mean that you will get celiac disease.

So looking at your genetic polymorphisms could help you rule out celiac disease, but not tell you if you have it.

***This information all comes with the disclaimer that you should always talk with your doctor for a diagnosis.  Seriously!  Your doctor can order tests such as blood test or a small intestine biopsy to accurately diagnose or rule out celiac disease.***

A little science background…
So what is an HLA serotype?  Serotypes, discovered in 1933 by Rebecca Lancefield, are variations within species (bacteria and viruses) or among immune cells of people.  The typing is based on their cell surface antigens.   In humans, the human leukocyte antigen (HLA) determines the serotype; the HLA serotype is sometimes used in determining transplant matches.

One HLA type is HLA-DQ, which is a protein found on antigen presenting cells.  DQ is involved in the immune system through stimulating T-cells which then signal B-cells to produce antibodies.  The HLA-DQ recognizes foreign antigens from pathogens, but it also recognizes common self-antigens.  This is where the problem begins when the HLA-DQ loses it tolerance to self-proteins, triggering autoimmune diseases such as celiac, lupus, and type 1 diabetes.

SNP’s involved in Celiac:
Approximately 90-95% of celiac patients have HLA-DQ2.5, which can be determined from the SNP rs2187668 (T). [ref] The SNP is fairly common and found in about 25% of European caucasians.  So, again, having the SNP does not mean that you have Celiac, only that it is possible for you to have it or develop it at some point.

Check your 23andMe results for rs2187668:

  • CC: wildtype (no risk for Celiac unless HLA-DQ8)
  • CT: one allele for HLA-DQ2.5 (Celiac disease is possible)
  • TT: two alleles for HLA-DQ2.5 (Celiac disease is possible)

HLA-DQ8 alone is found in about 5-10% percentage of Celiac patients.

Check your 23andMe results for rs7454108:

  • CC: two alleles for HLA-DQ8 (Celiac disease is possible)
  • CT: one allele for HLA-DQ8 (Celiac disease is possible)
  • TT: wildtype

The SNPs above cover most people with celiac, but not quite all.  The HLA-DQ2.2 variant, found by itself in a very small percentage of celiac cases, can not be completely determined with the current version of 23andMe data.

Other genes involved in celiac….
Since about 1% of those with European ancestry  receive a diagnosis of celiac while 30% have the HLA types for it, something else must also be involved in a person’s risk for celiac.  Recent studies have looked into other genetic traits, as well as environmental aspects such as duration of breastfeeding, timing of the introduction of wheat to an infant’s diet, and types of bacteria in the gut.

Below are a few of the current studies on possible genetic causes of celiac disease.

A large 2011 study looked at genetic differences between over 12,000 individuals with celiac compared to approximately the same number without celiac disease.  This study identified quite a few polymorphisms that influence the risk of celiac disease.  Look at Table 2 in the study to see the SNPs and their risk.

A sibling study from 2011 found three SNPs to be significant in increasing the risk of celiac disease among siblings where one sibling had been diagnosed with celiac.  The A-alleles of rs1464510, rs842647, and rs2816316 increased the risk of celiac disease in siblings in conjunction with the HLA type.  See Table 2 in the study.

Check your 23andMe results for rs1464510:

  • AA: increased risk of celiac
  • AC: increased risk of celiac
  • CC: no increased risk
Check your 23andMe results for rs842647:

  • AA: increased risk of celiac
  • AG: increased risk of celiac
  • GG: no increased risk
Check your 23andMe results for rs2816316:

  • AA: increased risk of celiac
  • AC: increased risk of celiac
  • CC: no increased risk

An August 2015 meta analysis showed that a T-allele (23andMe orientation) on rs917997  increased the risk of celiac disease by 5%, The same study showed that a T-allele on rs6441961 increased the risk of celiac by 6%.

Check your 23andMe results for rs917997:

  • TT: increased risk of celiac
  • CT: increased risk of celiac
  • CC: no increased risk
Check your 23andMe results for rs6441961:

  • TT: increased risk of celiac
  • CT: increased risk of celiac
  • CC: no increased risk

A Finish study from 2012 found that FUT2 non-secretors are at an increased risk of celiac disease.  The AA genotype for rs601338 determines a non-secretor.  The odds ratio for non-secretors for celiac is 1.28.  (Non-secretors are also resistant to the Norovirus.)

Check your 23andMe results for rs601338:

  • AA: increased risk of celiac
  • AG: no increased risk
  • GG: no increased risk

It seems that there is quite a bit of research going on right now into why one person with HLA DQ2.5 will have celiac disease while another won’t ever get it.  For anyone who is interested in further researching the topic, a PubMed search of “Celiac polymorphism” currently returns over 1800 results.  One other aspect that has recently been studied is the effect of the gut microbiome.   A 2015 paper in the journal, Nutrients, summed up the current research into the gut microbe connection.  Another 2015 article in the American Journal of Pathology is worth reading and tantalizing.    It is exciting to see how close we are to finally understanding this disease.

Comments 36

  • […] Celiac disease (coeliac) is an autoimmune disorder in which gluten causes damage to the villi in the small intestines. Gluten is a protein found in wheat, barley, rye, and spelt. There are many re…  […]

  • […] read through the Genetic SNPs for Celiac Disease first to determine if you are a carrier of HLA DQ2.5 or HLA […]

  • That’s interesting. I just had my genotype / DNA analysed and have come back with 35 SNPs associated with coeliac disease; top of the list is 2187668-A. I have always tested negative for CD but adopted a gluten-free diet to see if it would ease what was thought to be IBS. It did, but more than that it stopped the systemic, whole-body itching that had been plaguing me for at least 20 years. And now I find that if I eat anything with wheat in it (haven’t been able to try with rye and barley) I start itching again. I do wonder now it was CD, just with skin problems (invisible HD?) as the main symptom…

  • […] Celiac disease (coeliac) is an autoimmune disorder in which gluten causes damage to the villi in the small intestines. Gluten is a protein found in wheat, barley, rye, and spelt. There are many re…  […]

  • Thank you for this interesting post.

    Will these values be the same in data from ancestrydna?

    • Thanks for checking out my blog! The rs numbers will be the same no matter where you get your dna information, but you may find that some of the snps that I’ve listed aren’t sequences by ancestrydna.

  • Very helpful post; thanks so much for the great information! I found your blog this morning and have devoted my Saturday morning coffee time to exploring it. :-)

  • […] with what ever topic is of interest to you.  Figure out if you are genetically able to get celiac disease.  Check out your hemochromatosis genes.  Or jump into the methylation cycle and figure out if you […]

  • Hi! Thank you for the article. So I need some basic spelling out here. I seem to be rs2187668 CC and the other-TT-so no celiac-but thrones down below: rs1464510 and rs842647 says I have a risk-so do I or not because I did not meet the criteria for the first 2? Thank you!

  • Your article left me confused. You are saying that rs2187668 is HLA-DQ2, yes? I have CC for rs2187668, but was HLACEL tested and was found to be positive for HLA DQA1 05 and HLA DQB1 02 which are HLA-DQ2. Why the difference, if rs2187668 is HLA-DQ2?

  • I was considering genetic testing to try and rule out dermatitis herpetiformis but now I am being told that there is no correlation. My biopsies and labs for celiac have come back negative. Doctors still think it could be the dermatitis herpetiformis. If I do not have the celiac gene, can DH be completely ruled out?

    • Hi Donna – You are right that you can get DH without celiac disease, but I’m not sure that you can rule out DH completely just by genetic or HLA testing. According to this study (, 80 – 90% of dermatitis herpetiformis cases are the same HLA type (HLA-DQ2.5, DQ8) as celiac disease, but that would still leave 10-20% of cases that don’t fit in those genotypes.
      Good luck on getting this figured out.

  • I have TC for rs2187668 and TT for rs7454108. I understand I am at risk with the first finding, but what does the wildcard mean on the HLA DQ8 mean?


    • If you have the wildtype for HLA-DQ8 that just means that you have the normal type that the majority of people have. It doesn’t add (or subtract) from the risk for Celiac.
      Basically, you need to have either of the HLA types to be able to have Celiac, along with other environmental requirements that are still being studied. There have been a couple of recent studies on the changes to the gut microbiome in celiac patients as well as links to other autoimmune conditions such as Hashimoto’s thyroiditis.

      • Hey.. not sure if youre still checking here… I was dx’ed with celiacs about 16 years ago (based on anti gliadin antibodies) back when it wasn’t glamorous and there was hardly any marketing for products, and have been GF since. If ever I get cross contamination I def get sick. Since being GF I no longer need meds for Hashis. And yet I don’t have the main markers for Celiacs. I did have 3 of the others you mention (rs842647, rs2816316, rs6441961). I guess my fear is that when they begin to have treatments (beyond maintain a gf diet) folks like me will be left out due to lack of clear cut evidence. I’m horrible with “the maths” so I guess I’m not really understanding these 5% increased risk genes. Do they simply point to the possibility that someone can develop Celiacs with these genes? I guess I’m just bummed not to have the smoking bullet (besides those antibodies). :/

        • So not having any of the HLA types for Celiac makes it unlikely- but not impossible – to have Celiac. There is certainly a possibility of errors in the DNA testing and there could be other rare exceptions of HLA types. The other genes listed only add to the mathematical risk of Celiac if you have the HLA type.

          The flip side is that the anti-gliadin antibodies test is also no longer considered reliable for diagnosing Celiac disease. Here is an article on it:

          If at some point they have a treatment beyond diet, you could always have HLA specific blood typing done to make sure that the genetic test is right. But all of this, whichever test is right, is probably a moot point for you if you feel better on a gluten-free diet and it has cured your Hashimoto’s! Even if they come up for a treatment for Celiac, you may still want to stay on the gluten-free diet just for Hashi’s.

          • For sure I definitely think wheat/gluten is my kryptonite and certainly I avoid it. After going GF I had some moments when I thought: Maybe I don’t have Celiacs (because who would want to have it?) but then I get cross contaminated and I’m back to square one. I guess I’m just annoyed as I thought it was the one thing my DNA would make clear. But then again, according to 23&me I don’t have the genes for red hair or freckles or cleft chin and my hair shouldnt be curly… I guess they haven’t got it all worked out just yet…. I have never heard of the HLA specific blood tests, but maybe I’ll have my functional med doc look into it… if its not too costly it would be nice to know. Thanks so much for the suggestion! You rock!

  • […] friends, we now have our genetic test results from 23andMe.  And I do have two gene variants that are linked to Celiac disease.  I’m not at all surprised that I also have the MTHFR […]

  • Thank you so much for posting the SNP’s to search 23andMe. I have been searching everywhere for this information since it is why I took the test. I found out too late that they will be offering this information, but haven’t rolled it out yet. I stopped eating gluten due to a low carb diet and a host of symptoms disappeared. I am investigating other issues that might cause this but I wanted to check the genes for Celiac first. I do not have the first two but do have 3 of the (more minor?) SNP’s. Not exactly a smoking gun, but sounds like enough to continue testing anyway.

    • Hi –
      If you don’t have the HLA DQ2.5 or HLA DQ8 genes, it is unlikely that you will have celiac. I say “unlikely” because there is always a chance that 23andMe has your information incorrect and also because most studies list those two HLA types as covering about 99% of cases — leaving a little 1% wiggle room.
      That said, if you feel better without gluten, that is reason enough not to eat it. Whether you label it non-celiac gluten sensitivity or name it anything else, if you feel better without gluten, your body is telling something. It is a lot easier, though, to eat gluten-free if you don’t have to worry as much about cross-contamination as you would with celiac disease. And perhaps, at some point, you can add it back into your diet. I’ve been able to eliminate the gut-related gluten issues for myself by adding in bifidobacteria probiotics, but gluten still reliably gives me a migraine, which leaves it off my menu!
      Thanks for reading my blog!

      • Good to know! With a 1% chance I won’t bother with a gluten challenge then. It didn’t sound like a fun idea since so many issues went away after I went GF. I don’t know if cross contamination is an issue for me but I suspect it is mild if anything. Something is still causing joint swelling and inflammatory response but nowhere near where it was, so it could be another problem altogether. Migraines and digestive issues are down to a dull roar so I certainly don’t have the CC issues I’ve seen described by some. I have started probiotics-hopefully they will help! Thanks again for all the information.

        • LA, just because you don’t carry the genetic markers for celiac doesn’t at all mean you can’t get a general immune response to gluten. It just means the autoimmune disease called celiac isn’t likely. Autoimmune disease and immune response are different. In one, your body attacks the foreign invaders and in the other it attacks itself. If you do any research on the gut biome and leaky gut, you’ll find you can easily become ‘allergic’ or ‘sensitive’ to any foreign protein that comes into contact with the cells responsible for collecting and cataloging substances they believe to be dangerous intruders. This can happen through your skin, lungs, internal organs, etc. — basically any way bacteria or viruses could get in, your immune cells will be on the alert. If you do better off of gluten, it’s likely that gluten is contributing to an inflammatory state in your body. Inflammation is a major contributor to pain and disease and not to be taken lightly simply because your DNA doesn’t lean towards the autoimmune version of gluten sensitivity. I hope that makes sense.

  • Per genetic testing that a gastroenterologist did for me in 2012, I have one copy of HLA-DQ2.5 and one copy of HLA-DQ2.2. My results state that I’m homozygous for HLA-DQ2, and my risk for celiac disease is “Extremely High.” My testing was done through Prometheus Therapeutics & Diagnostics. The test was called Prometheus Celiac Plus and also included serological markers (the latter of which were negative; I gave up gluten in 2006 before getting tested for celiac disease).

    I was surprised to see my 23andme report state that I have a “slightly increased risk” for celiac disease. From the text in my report, I gathered that they didn’t check for HLA-DQ2.2. That’s unfortunate and is misleading for people like me (and for people who have one or two copies of HLA-DQ2.2—without HLA-DQ2.5).

    So I’ve been poking around online about this matter and found your great article. Very well done! Thank you!

    Unfortunately, for those six other SNPs you describe, I’ve got increased risk for celiac disease with all of them. I’m heterozygous on four of them and homozygous on two of them.

    To this day, I still have digestive woes. I continue to wonder if I actually have celiac disease but am unwilling to do a weeks-long gluten challenge. I’m happy to hear a new test is in the works for people like me, where I’d eat gluten once and then get a simple blood test. Based on the result, I could finally have my answer. I’m willing to eat gluten once!

  • Thanks to this article, I was able to discover I have HLA-DQ8 risk (TC) for celiac but not HLA-DQ2.5 risk (CC). Out of the six markers listed above that increase risk, three were not sequenced by my DNA testing company and three carried increased risk for celiac. I am wondering how to look for HLA-DQ2.2.

    Can you tell me if that is something I can look for in my raw data or if I absolutely have to go through Prometheus’s Celiac assessment?

    I have been gluten-free for over two years and am very sensitive to even food made on equipment with gluten/wheat/barley/rye. I REALLY do not want to go on the ‘gluten challenge’ — pain & suffering for a month, tummy trouble, and extreme mood changes that make life a living hell for me and everyone around me. But somehow having a dx of celiac might make more sense to those around me and help with people believing I have something ‘real’ and not made up…

    • I totally understand the concept of trying to convince people it is real and not made up :-)

      If you have newer 23andMe data (v.5 — since August 2017), you can see your HLA DQ2.2 haplotype. Look for rs2395182 (T), rs7775228 (C), and rs4713586 (A).

      • Thanks so much for helping me find those markers. I have TT on rs2395182, TT on rs7775228 (where you said it should be C for the 2.2), and AA on rs4713586. So that’s two out of three. Does that mean I don’t have the 2.2 haplotype? And do you know anything about finding if you have the 7.5 or 2.5trans? It turns out my dad has BOTH the 2.5 and the 8, my grandpa has only the 8, and my mother’s sister has none of the markers for risk. So I think my risk comes from my dad’s side.

      • Also, I just found SNPedia’s website with a tidy little list of celiac risk markers. I am going down, checking if I have or not. For instance, rs6822844 I have GG — G carries risk and T reduces risk. Is this a source you might recommend or is there a better one to do my research with?

        • Snpedia is a great source and usually fairly accurate. It is curated by individuals (like wikipedia), so there is a possibility of errors, but for the most part, it is very useful.

          With just two out of the three SNPs for HLA DQ2.2, it means that you don’t have that HLA type.

          Here is a good study that goes into the SNPs that are associated with the HLA types for celiac: It lists HLA DQ7 as rs4639334, which isn’t in 23andMe. I’m not sure about HLA DQ7.5 or 2.5trans.

          Isn’t it cool to be able to trace it through several generations :-)

          • Yeah, AncestryDNA doesn’t yet sequence rs4639334 either (that’s who I did my test with, for genealogical reasons). I’ve definitely got HLA-DQ8 single haplotype though… I guess I can try to get the Prometheus test for all the sequencing to prove it to my doctor or just call it good with what I’ve got. :/ Thanks for the link and your super fast in-depth responses.

  • I’m wondering what it means if you get a wildtype for both the HLA-DQ2.5 and the HLA-DQ8?

    • Hi, If you have wildtype for both HLA-DQ2.5 and HLA-DQ8 it means that it is unlikely that you would have celiac disease. There is always the possibility of errors in the DNA testing, though, so you can’t rely on it completely.
      Thanks for reading!

  • My daughter’s test through Prometheus showed that she was ‘very high risk’ for celiac with a 2.5 trans. I did a lot of research and ended up paying out of pocket to have my whole family tested.

    Turns out, my husband has 2.2 and 2.5
    and I have double 7.5.

    1 child has 2.5 and 7.5
    4 of our children have 2.2 and 7.5
    In this situation, the 2 (from the 2.2) and the .5 (from the 7.5) can become a 2.5 by transition. Unfortunately this is often overlooked because it is not a naturally occurring 2.5. 23andme doesn’t look for it and articles don’t always mention it. (2.2 was found by promethease when I uploaded the 23andme raw data, but no mention of the 7.5.)

    For the record, none of my kids are confirmed celiac, only the above daughter, who has severe health issues, was tested and she was already gluten free before testing was done. All of my kids are gluten free now. Two of my kids with the 2.5trans have enough digestive issues to swear off gluten for good, they won’t cheat. The other kids cheat now and then without much consequence, even the natural 2.5

Leave a Reply

Your email address will not be published. Required fields are marked *